When a coding agent fails, the instinct is to examine the model or the prompt. A new paper argues that this instinct misses half the problem: real-world agent failures split roughly evenly between tool-use accuracy (how well the agent invokes a tool) and intrinsic tool accuracy (whether the tool itself

Training effective software engineering agents requires large volumes of high-quality task trajectories, and generating them at scale is expensive. A new paper extends the "Less-Is-More" (LIMO) hypothesis — which showed in mathematical reasoning that a small number of excellent examples dramatically outperforms large volumes of mediocre ones — to the

The largest real-world study of AI-assisted coding in IDE-native settings to date has landed, and its findings challenge some of the most common assumptions about how developers actually use these tools. Researchers analyzed 74,998 developer messages across 11,579 chat sessions — all drawn from real sessions committed to public

A new large-scale empirical study has done something the agentic coding field has been waiting for: it went outside the benchmarks entirely and examined 110,000 real open-source pull requests from five production coding agents — OpenAI Codex, Claude Code, GitHub Copilot, Google Jules, and Devin — to measure what actually happens

The dominant paradigm in reasoning-augmented code generation has a specific shape: think first, then write. Chain-of-thought reasoning happens before the code begins, the model works out the problem, and then execution starts. A new paper from Xue Jiang, Tianyu Zhang, Ge Li, and colleagues challenges whether this front-loaded design is

Logging is supposed to be the signal that tells you what went wrong. For LLM-based coding agents, it's largely broken — and a new paper from Xin Wang and colleagues explains precisely why. Current automatic logging systems are trained to produce logs that look like developer-written logs, optimized for

There's a gap in the agentic coding conversation that most practitioners hit quickly: single-agent mode works for demos, breaks at production scale, but the jump to "use a full framework" feels like overkill for a team of three. Christopher Montes hit that ceiling with Claude Code

Bug attribution — figuring out which commit introduced a defect — has been one of software engineering's most stubborn unsolved problems. The SZZ algorithm, which just received the 2026 ACM SIGSOFT Impact Award, was published twenty years ago and still underpins most fault-attribution research today. The best traditional approach published